×

On a strategy to develop robust and simple tariffs from motor vehicle insurance data. (English) Zbl 1097.62112

The goals of this paper are twofold: we describe common features in data sets from motor vehicle insurance companies and we investigate a general strategy which exploits the knowledge of such features. The results of the strategy are a basis to develop insurance tariffs. We use a nonparametric approach based on a combination of kernel logistic regression and \(\varepsilon\)-support vector regression which both have good robustness properties. The strategy is applied to a data set from motor vehicle insurance companies.

MSC:

62P05 Applications of statistics to actuarial sciences and financial mathematics
62G08 Nonparametric regression and quantile regression
62G35 Nonparametric robustness
91B30 Risk theory, insurance (MSC2010)

Software:

e1071
PDFBibTeX XMLCite
Full Text: DOI Link

References:

[1] Bartlett, P.L., Tewari, A. Sparseness vs Estimating Conditional Probabilities: Some Asymptotic Results. Preprint, University of California, Berkeley, 2004 · Zbl 1078.68684
[2] Celebrián, A.C., Denuit, M., Lambert, P. Generalized Pareto Fit to the Society of Actuaries’ Large Claims Database. North American Actuarial Journal, 7: 18–36 (2003) · Zbl 1084.62108
[3] Cherkassky, V., Ma, Y. Practical selection of SVM parameters and noise estimation for SVM regression. Neural Networks, 17: 113–126 (2004) · Zbl 1075.68632 · doi:10.1016/S0893-6080(03)00169-2
[4] Christmann, A., Fischer, P., Joachims, T. Comparison between various regression depth methods and the support vector machine to approximate the minimum number of misclassi.cations. Computational Statistics, 17: 273–287 (2002) · Zbl 1010.62054 · doi:10.1007/s001800200106
[5] Christmann, A., Rousseeuw, P.J. Measuring overlap in logistic regression. Computational Statistics and Data Analysis, 37: 65–75 (2001) · Zbl 1051.62065 · doi:10.1016/S0167-9473(00)00063-3
[6] Christmann, A., Steinwart, I. On robust properties of convex risk minimization methods for pattern recognition. Journal of Machine Learning Research, 5: 1007–1034 (2004a) · Zbl 1222.68348
[7] Christmann, A., Steinwart, I. Consistency and robustness of kernel based regression. University of Dortmund, SFB-475, TR-01/05 Submitted, 2005 · Zbl 1129.62031
[8] Embrechts, P., Klüppelberg, C., Mikosch, T. Modelling Extreme Events for Insurance and Finance. Springer-Verlag, Berlin, 1997 · Zbl 0873.62116
[9] Hastie, T., Tibshirani, R. Classification by pairwise coupling. Annals of Statistics, 26: 451–471 (1998) · Zbl 0932.62071 · doi:10.1214/aos/1028144844
[10] Keerthi, S.S., Duan, K., Shevade, S.K., Poo, A.N. A fast dual algorithm for kernel logistic regression. In Machine Learning: Proceedings of the Ninetheenth International Conference, Kaufmann, San Francisco, 299–306 2004 · Zbl 1085.68127
[11] Leisch, F. et al. R package e1071. http://cran.r-project.org, 2003
[12] Nelder, J.A., Mead, R. A simplex algorithm for function minimization. Computer Journal, 7: 308–313 (1965) · Zbl 0229.65053
[13] Rousseeuw, P.J., Christmann, A. Robustness against separation and outliers in logistic regression. Computational Statistics & Data Analysis, 43: 315–332 (2003) · Zbl 1429.62325 · doi:10.1016/S0167-9473(02)00304-3
[14] Rüping, S. myKLR - kernel logistic regression. Department of Computer Science, University of Dortmund. http://www-ai.cs.uni-dortmund.de/SOFTWARE, 2003
[15] Schölkopf, B., Smola, A. Learning with Kernels. Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, Cambridge, Massachusetts, 2002
[16] Steinwart, I. On the Influence of the Kernel on the Consistency of Support Vector Machines. Journal of Machine Learning Research, 2: 67–93 (2001) · Zbl 1009.68143 · doi:10.1162/153244302760185252
[17] Vapnik, V. Statistical Learning Theory. John Wiley & Sons, New York, 1998 · Zbl 0935.62007
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.